code

library(rvest)
library(tidyverse)
library(data.table)
library(tidytext)
library(stringr)
library(ggplot2)
library(ggrepel)
library(shiny)
library(data.table)
library(leaflet)
library(plotly)
library(corrplot)
library(lubridate)
library(readr)
library(reshape2)
library(ggmap)
library(leaflet)

ggmap::register_google(key = "AIzaSyCWJ8DwoFBs1kYzM82SSA-XAvBDXT8YJRA")

bay <- ggmap(get_googlemap(center = c(longitude = -122, latitude = 37.9),
                    zoom = 9, scale = 1, 
                    maptype = "roadmap",
                    color = "color"))

  # leaflet() %>%
  # addTiles() %>%
  # setView(lng = -122,lat = 38, zoom = 7.5) %>%
  # addMarkers(lat = boba.data$lat,
  #            lng = boba.data$long, #west is negative, east is positive
  #            # label = name,
  #            popup = boba.data$name) #textbox for more description (long text that would take long to upload)


boba.data <- read_csv("boba_bay_area.csv")

city.data <- boba.data %>%
  group_by(city) %>%
  summarize(mean =  mean(rating), 
            mean.lat = mean(lat),
            mean.lng = mean(long))

colnames(city.data) <- c("City", "Average Rating", "Latitude", "Longitude")

city.rating <- bay + 
  geom_point(data = city.data,
             aes(col = class, x = Longitude, y = Latitude)) +
  guides(alpha = FALSE) +
  ggtitle("Average Yelp Rating by City") +
  theme(axis.title.x = element_blank(),
        axis.title.y = element_blank()) +
  labs(color= "Yelp ratings")

# city.rating.plotly <- ggplotly(city.rating, tooltip = c("City", "Average Rating"))

# boba.map <- bay + 
#   geom_point(data = boba.data,
#              aes(col = factor(rating), x = long, y = lat), size = 2) +
#   guides(alpha = FALSE) +
#   ggtitle("Boba Bae in Your Area") +
#   theme(plot.title = element_text(color = "black", size = 14, face = "bold.italic"),
#         axis.title.x = element_blank(),
#         axis.title.y = element_blank()) +
#   labs(color= "Yelp ratings")

## RANKING BY CITY
city.data$`Average Rating` <- (city.data$`Average Rating`) - 2.5
city.data <- city.data %>%
  arrange(-city.data$`Average Rating`)

city.data <- city.data %>%
  mutate(class = ifelse(city.data$`Average Rating`>1.7, 
                        "Top 10%",
                        ifelse(city.data$`Average Rating`>1.456, 
                               "Top 20%",
                               "Rest")))

city.data$class <- factor(city.data$class, levels = c("Top 10%", "Top 20%", "Rest"))

# city.data %>%
#   ggplot(aes(x = reorder(city.data$City, city.data$`Average Rating`), y = city.data$`Average Rating`)) +
#   geom_bar(aes(fill = class), stat = "identity") +
#   xlab("City") +
#   ylab("Average Rating") +
#   scale_y_continuous(labels = c("2.5", "3", "3.5", "4", "4.5")) +
#   coord_flip()+
#   ggtitle("Average Rating by City (including ties)")

## RANKING BY NAME
store.data <- boba.data %>%
  group_by(name) %>%
  summarize(mean =  mean(rating)) %>%
  arrange(mean)

store.data <- store.data %>%
  mutate(class = ifelse(store.data$mean == 2, 
                        "Lowest",
                        ifelse(store.data$mean == 2.5, 
                               "Second to the Lowest",
                               ifelse(store.data$mean == 4.5,
                                      "Second to the Highest",
                                      "Highest"))))
store.data.low <- store.data %>%
  filter(mean <= 2.5)

store.data.high <- store.data %>%
  filter(mean >= 4.5)

store.data.low.high <- rbind(store.data.low, store.data.high)

store.joined <- merge(store.data.low.high, boba.data[,c(3,5,6,7,8)], by = "name")

# city.data %>%
#   ggplot(aes(x = reorder(city.data$City, city.data$`Average Rating`), y = city.data$`Average Rating`)) +
#   geom_bar(aes(fill = class), stat = "identity") +
#   xlab("City") +
#   ylab("Average Rating") +
#   scale_y_continuous(labels = c("2.5", "3", "3.5", "4", "4.5")) +
#   coord_flip() +
#   ggtitle("Average Rating by City (including ties)")
# 
# bay + 
#   geom_point(data = city.data,
#              aes(col = factor(class), x = city.data$Longitude, y = city.data$Latitude), size = 2) +
#   guides(alpha = FALSE) +
#   ggtitle("Ranking by City") +
#   theme(plot.title = element_text(color = "black", size = 14, face = "bold.italic"),
#         axis.title.x = element_blank(),
#         axis.title.y = element_blank()) +
#   labs(color= "Class")
  
ui <- fluidPage(
  titlePanel("Where's My Boba Bae?"),
  sidebarLayout(
    sidebarPanel(
      tags$head(
        tags$style("description {white-space: nowrap;}")),
      fluidRow(textOutput(outputId = "intro"))
    ),
    mainPanel(
      tabsetPanel(
        tabPanel("The snapshot",
                 verbatimTextOutput(outputId = "snapshot"),
                 plotlyOutput(outputId = "whole.map")),
        tabPanel("Ranking by Cities",
                 plotOutput(outputId = "city"),
                 plotlyOutput(outputId = "store.map"), 
                 radioButtons(inputId = "class",
                             label = "Choose your preferred city ranking level",
                             choices = c("Top 10%", "Top 20%", "Rest"),
                             selected = city.data$class[1]),
                 leafletOutput(outputId = "city.map"),
                 tableOutput("levelData"),
                 br()),
        tabPanel("Ranking by Stores",
                 textOutput(outputId = "store.title"),
                 plotOutput(outputId = "store.low"),
                 plotOutput(outputId = "store.high"),
                 plotlyOutput(outputId = "store.ranking"),
                 radioButtons(inputId = "store",
                              label = "Choose your preferred store ranking level",
                              choices = c("Lowest", "Second to the Lowest", "Second to the Highest", "Highest"),
                              selected = store.joined$class[1]),
                 plotlyOutput(outputId = "store.select")),
        tabPanel("Boba Bae Options", 
                 selectInput(inputId = "select",
                             label = "Find your potential boba bae in your preferred city!",
                             choices = boba.data$city,
                             selected = boba.data$city[1]),
                 leafletOutput(outputId = "boba.map"),
                 textOutput(outputId = "text"),
                 br())
      )
    )
  )
)

server <- function(input, output, session) {
  output$intro <- renderText(
    paste("You 1) have never tried boba before; 2) have just recognized the contagious presence of boba, but don't know which one to try (because there are just so many of them!); or 3) are a boba enthusiast and are willing to try more but need help finding them efficiently. Worry no more. This interactive app helps you find your Boba Bae in a pool of 603 boba stores located in the Bay Area.")
  )
  
  output$whole.map <- renderPlotly({
    
    bay.whole.map <- bay + 
      geom_point(data = boba.data,
                 aes(col = factor(rating), x = long, y = lat), size = 1) +
      guides(alpha = FALSE) +
      ggtitle("Boba Bae in Your Area") +
      theme(plot.title = element_text(color = "black", size = 14, face = "bold.italic"),
            axis.title.x = element_blank(),
            axis.title.y = element_blank()) +
      labs(color= "Yelp ratings")
    
    mytext = paste("Store =", boba.data$name, "\n", "Yelp Rating =", boba.data$rating)
    
    pp = plotly_build(bay.whole.map)
    style(pp, text = mytext, hoverinfo= "text")
  })
  
  output$snapshot <- renderText({
    paste("This maps shows all the 603 boba stores, ranked by color. Though the greater San Jose area is the most populated region (and note how the high-ranked ones are usually located there), there are other options scattered around in the Bay Area.")
  })
  
  output$boba.map <- renderLeaflet({
    input.city <- boba.data %>%
      mutate(is.selected.city = 
               ifelse(boba.data$city == input$select,
                      "city",
                      "not.city")) %>%
      filter(is.selected.city == "city")
  
    input.city %>%
      leaflet() %>%
      # setView(lng = boba.data$long[1],lat = boba.data$lat[1],zoom = 8) %>%
      addTiles() %>%
      addCircleMarkers(lat = ~lat,
                       lng = ~long, 
                       clusterOptions = markerClusterOptions(),
                       popup = paste(input.city$name, "<br>",
                                     "rating:", input.city$rating, "<br>")) #zipcode also needs to be a string
    
  })

  output$city.rating <- renderPlot({
    
    city.data %>%
      ggplot(aes(x = "City", y = "Average Rating")) %>%
      geom_bar(stat = "identity")
    
  })
  
  output$city <- renderPlot({
    city.data %>%
      ggplot(aes(x = reorder(city.data$City, city.data$`Average Rating`), y = city.data$`Average Rating`)) +
      geom_bar(aes(fill = class), stat = "identity") +
      xlab("City") +
      ylab("Average Rating") +
      scale_y_continuous(labels = c("2.5", "3", "3.5", "4", "4.5")) +
      coord_flip() +
      ggtitle("Average Rating by City (including ties)")
  })
 
  output$store.title <- renderText({
    paste("Average Ranking by Stores")
    
  })
  output$store.low <- renderPlot({
    store.data.low %>%
      ggplot(aes(x = reorder(store.data.low$name, store.data.low$mean), y = store.data.low$mean)) +
      geom_bar(aes(fill = as.factor(class)), stat = "identity") +
      xlab("Store") +
      ylab("Average Rating") +
      coord_flip()+
      ggtitle("All-time low") +
      scale_fill_discrete("Class")
  }) 
  
  output$store.high <- renderPlot({
    store.data.high %>%
      ggplot(aes(x = reorder(store.data.high$name, store.data.high$mean), y = store.data.high$mean)) +
      geom_bar(aes(fill = as.factor(class)), stat = "identity") +
      xlab("Store") +
      ylab("Average Rating") +
      coord_flip()+
      ggtitle("All-time high") +
      scale_fill_discrete("Class")
  })
  
  output$store.ranking <- renderPlotly({
    store.level <- bay + 
      geom_point(data = store.joined,
                 aes(col = factor(class), x = store.joined$long, y = store.joined$lat), size = 2) +
      guides(alpha = FALSE) +
      ggtitle("Overview of Stores by Rankings") +
      theme(plot.title = element_text(color = "black", size = 14, face = "bold.italic"),
            axis.title.x = element_blank(),
            axis.title.y = element_blank()) +
      labs(color = "class")
    
    store.text = paste("Store =", store.joined$name, "\n", "Average Rating =", store.joined$mean)
    
    store.plotly = plotly_build(store.level)
    style(store.plotly, text = store.text, hoverinfo= "text")
  })
  
  output$store.select <- renderPlotly({
    store.data <- store.joined %>%
      filter(class == input$store)
    
    store.level <- bay + 
      geom_point(data = store.data,
                 aes(col = factor(class), x = store.data$long, y = store.data$lat), size = 2) +
      guides(alpha = FALSE) +
      ggtitle("Stores by Preferred Rankings") +
      theme(plot.title = element_text(color = "black", size = 10, face = "italic"),
            axis.title.x = element_blank(),
            axis.title.y = element_blank()) +
      theme(legend.position = "none")
    
    store.text = paste("Store =", store.data$name, "\n", "Average Rating =", store.data$mean)
    
    store.plotly = plotly_build(store.level)
    style(store.plotly, text = store.text, hoverinfo= "text")
  })
  
  output$city.map <- renderLeaflet({
    
    input.class <- city.data %>%
      mutate(is.selected.class = 
               ifelse(city.data$class == input$class,
                      "class", 
                      "not.class")) %>%
      filter(is.selected.class == "class")
    
    input.class %>%
      leaflet() %>%
      addTiles() %>%
      # setView(lng = -122,lat = 37.9, zoom = 8) %>%
      addCircleMarkers(lat = ~Latitude,
                       lng = ~Longitude, 
                       clusterOptions = markerClusterOptions(),
                       popup = paste(input.class$class, "<br>",
                                     "City: ", input.class$City, "<br>")) #zipcode also needs to be a string
    
  })
  
  output$store.map <- renderPlotly({
    city.data.adjusted <- city.data %>%
      mutate(adjusted.rating = city.data$`Average Rating` + 2.5)
    
    city.data.adjusted$adjusted.rating <- round(city.data.adjusted$adjusted.rating, digit = 2)

    snapshot.level <- bay + 
      geom_point(data = city.data.adjusted,
                 aes(col = factor(class), x = city.data.adjusted$Longitude, y = city.data.adjusted$Latitude), size = 2) +
      guides(alpha = FALSE) +
      ggtitle("Snapshot of city rankings") +
      theme(plot.title = element_text(color = "black", size = 10, face = "italic"),
            axis.title.x = element_blank(),
            axis.title.y = element_blank()) +
      labs(color= "Class")
    
    snapshot.text = paste("City =", city.data.adjusted$City, "\n", "Average Rating =", city.data.adjusted$adjusted.rating)
    
    snapshot.plotly = plotly_build(snapshot.level)
    style(snapshot.plotly, text = snapshot.text, hoverinfo= "text")
  })
  
  output$text <- renderText({
    city.ranking <- city.data %>%
      filter(City == input$select) %>% #take data set and give only one row with state name = whatever the user clicks
      select(`Average Rating`) 
    
    city.ranking <- city.ranking + 2.5
    city.ranking <- round(city.ranking, digits = 2)
    
    paste("You have just clicked on", input$select,
          ", and that city's average rating of boba stores is", city.ranking, ". See 'Ranking by Cities' for detailed information about respective cities.")
  })
  
  output$levelData <- renderTable({
    city.level.filter <- city.data %>%
      mutate(level.filter <- city.data$`Average Rating` + 2.5)
    
   colnames(city.level.filter) <- c("City", "Adj. Average Rating", "Lat", "Long", "Class", "Average Rating")
    
    level.filter <- subset(city.level.filter[,c(1,6)], city.data$class == input$class)
  })
}

shinyApp(ui = ui, server = server)